Frequently Asked Questions

Thumba is an enterprise AI platform that converts requirements, user stories, workflows, and project documents into high-quality user stories, acceptance criteria, and test cases, while identifying gaps that cause rework and go-live delays.

Thumba is designed for teams involved in planning, building, and validating software delivery, including Product Managers, Business Analysts, Quality Analysts, Scrum Masters, and Engineering Managers. Each role uses Thumba to reduce ambiguity, eliminate rework, and ensure sprint-ready artefacts before development begins.

Thumba is domain-agnostic and is actively used across regulated and non-regulated industries such as Banking, Insurance, FinTech, Healthcare, Real Estate, Manufacturing, and Software Product companies. The models are designed to respect domain terminology, compliance constraints, and project-specific glossaries.

• AI User Story & Acceptance Criteria Generation — Converts requirements, epics, documents, and workflows into structured user stories and acceptance criteria.
• AI Test Scenario & Test Case Generation — Automatically generates test scenarios, test cases, test steps, and expected results for QA teams.
• User Story Gap Detection & Auto-Fix — Scans user stories for ambiguity, inconsistency, missing scenarios, and fixes them using industry standards.
• Workflow & Image-Based Generation— Converts Visio files and workflow images into executable user stories and test cases.

Yes. Thumba accepts project documents such as BRD, PRD, FSD, architecture diagrams, and supporting artefacts in DOCX and PDF formats. Uploading project artefacts significantly improves context, coverage, and accuracy of generated user stories and test cases.

The accuracy of Thumba’s output depends on the quality and completeness of input data. Customers have reported up to 95% accuracy in generated user stories and test cases when project documents and artefacts are included.

Thumba is built keeping enterprise grade privacy and data protection for all its users. All the artefacts fetched from your source ALM or uploaded documents are NOT STORED in Thumba, similarly all outputs are either downloaded into your local folder or the user can manually submit back to source project in the respective ALM. We retain zero user data or project information except for limited details to connect with your source project. All input and generated data are deleted once user navigates to different page or post their logout. Thumba does not use customer data for model training and follows enterprise-grade security and privacy practices.

Thumba is headquartered in UK, with teams spread across London, Trivandrum and Bangalore.

Unlike generic AI tools, Thumba is purpose-built for software delivery workflows. It combines requirement understanding, industry standards, consistency, and enterprise privacy to deliver predictable, production-ready outputs.

Poorly written user stories are a leading cause of sprint rework and delayed go-lives. Thumba scans user stories to identify ambiguity, contradictions, missing scenarios, and inconsistencies before development begins.

Yes. Thumba identifies gaps in user stories and acceptance criteria, such as ambiguity, missing scenarios, inconsistencies, and incomplete coverage and fixes them using industry best practices and your project documents. This ensures user stories are sprint-ready and require minimal manual review.

When user stories are fetched for scanning, Thumba performs a pre-scan sanity check to identify formation issues, grammatical errors, or unsupported characters. User stories flagged in yellow must be corrected before running the full scan to ensure accurate results.

These are the steps you can perform in scanning:

1. Integrate your project — Connect your Jira or Azure DevOps project. If user stories are in Excel, create an Excel-based project.
2. Fetch user stories — Navigate to Scanning and fetch user stories and acceptance criteria from the project.
3. Review pre-scan checks — Thumba performs a pre-scan to highlight basic issues that must be resolved before full scanning.
4. (Optional) Upload project glossary — Upload a glossary to ensure abbreviations and domain terms are interpreted correctly.
5. Run full scan — Select the user stories and initiate the full scan.
6. Review scan results — View gap reports directly in the UI, export them as Excel, or submit gaps back to Jira or Azure DevOps.
7. Upload supporting documents — Upload PRD, BRD, FSD, architecture diagrams, or related artefacts for better context and accuracy.
8. Fix gaps using Thumba — Select user stories with gaps and use Fix Gaps(s) to automatically fix them based on project artefacts and industry standards.
9. Edit and submit output — Review, edit if required, and download as Excel or submit back to Jira or Azure DevOps.
10. Watch step by step tutorial – How to detect gap in user stories & fix them using Thumba
    

This can happen for one of the following reasons:

1. No user stories exist in Jira or Azure DevOps
2. Incorrect column names in the uploaded Excel file
3. No action was taken within 30 minutes after fetching (you can re-fetch)
4. User stories are already in a Done or Fixed state

Fixed user stories are created as separate linked items in your project. The gap report is attached to each original user story for traceability and future review.

No. Thumba creates a new user story with the fixes applied and links it to the original user story. The original item remains unchanged.

Yes. You can upload user stories from an Excel file, provided the columns are correctly labelled according to Thumba’s supported format.

Currently, Thumba supports scanning of user stories and acceptance criteria only. Support for additional work item types may be added in future releases.

Generation from requirements enables teams to automatically convert raw requirements into structured delivery artefacts.

Using requirements from Jira, Confluence, Azure DevOps, or uploaded documents, Thumba can generate:

1. User stories
2. Acceptance criteria (standard and Gherkin formats)
3. Test scenarios
4. Test cases

This approach significantly reduces manual design effort for Business Analysts and QA teams, with customers reporting up to 85% reduction in design effort and up to 95% accuracy, depending on input quality and context provided.

You can upload or fetch requirements in the following ways:

• Fetch requirements or epics directly from Jira or Confluence
• Fetch requirements from Azure DevOps
• Upload requirements from an Excel file, ensuring the requirement column is clearly labelled.

Please verify the following:The requirements column is labelled exactly as “requirements”

• The requirements column contains data and is not empty
• The generation process may take a few seconds to several minutes, depending on volume and complexity.
• Do not refresh or navigate away during processing

If the issue persists, contact support at support.thumba@thumba.ai

Use the following checklist to troubleshoot:

1. Confirm that Jira or Azure DevOps project details are entered correctly during integration
2. Ensure all requirements are present in the project before initiating generation
3. Thumba fetches only requirements that are in Backlog or In Progress states

1. User Stories — Newly generated user stories are created within the same Jira project.
2. Test Scenarios and Test Cases — Generated test scenarios and test cases are stored in the configured test management tool, such as Zephyr or Xray, for that project.

If you have an existing set of user stories and acceptance criteria in your project, QA teams can generate Test Scenarios and Test Cases with the help of your project artefacts.

Yes, you can map each field that is user story and acceptance criteria in your project and fetch it into Thumba for generation of QA artefacts.

Yes, Thumba has two-way integration with your project in ALM. You can fetch user stories and acceptance criteria and submit them back to your source project in these ALM tools.

Thumba can accept both formats of user stories and acceptance criteria:

1. Gherkin Format (BDD)
2. Standard Format

You can convert use case diagrams or customer flows into detailed user stories with acceptance criteria for business and test cases for QA teams.

The workflow diagrams can be in: PNG, JPEG, or JPG or a MS Vision file.

There are no upper or lower limits to it and usually depends on the complexity of the image and project artefacts attached to it. However, if your credit limit is lower than 100, please reach out to your account administrator.

Yes, you can submit the generated responses (user stories or test cases) back to your preferred ALM tool whether Jira or Azure DevOps.

Watch how to create Jira Project here: https://youtu.be/xaSOcjyhIpU

Watch how to create Azure DevOps Project here: https://youtu.be/1Nvzq7zT0ig

Watch how to create Excel Project here: https://youtu.be/T3VLw4tDMJc

Following reasons could be the issue:

1. Your PAT (Personal Access Token) if its Azure Project or API tokens (Jira project) could have expired.
2. No project is available in your ALM tool based on the information provided.

You can link up to 50 projects from a single account for standard subscriptions. However, enterprise accounts could configure unlimited projects.

Thumba offers a 14-day free trial with 75 credits, allowing users to experience generation, scanning, and gap fixing without any payment details. Ideal for Pilot Projects. You can sign up for trial access here: https://www.thumba.ai/registration

Yes, you can purchase credits post your trial period. Visit “billing” from admin console to get full access to credit packages and pricing.

As an admin you can track credits from your billing section and bottom of navigation bar. As a user you can keep account of credits at the bottom of navigation bar.

You can find distribution of credit consumption when hovered on credits on the navigation bar. Please refer to the Thumba application for a detailed credit breakdown image.

There are no differential credit charges with or without project or supporting documents upload. The credits are deducted based on the number of artefacts generated and is independent of any document uploads. For example, if you have generated 5 user stories and 10 test cases with the help of BRD, the credit would be charged only for the output and not for BRD.

The credits are consumed at the scanning stage and it is based on the number of user stories you scan, irrespective of whether a user story and acceptance criteria has gaps or not. There are no separate credit deductions when Fix Gaps(s) is utilised for fixing gaps identified by Thumba in your user stories.

You can add users to Thumba from “user management” in navbar. You can configure their feature access and allocate credits based on their role and upcoming sprint needs.

Your subscription is valid for 3-months, after which you need to add minimal credits to continue generation and scanning.

Enterprise customers can have custom credit plans, unlimited project and user accounts. Please write to us: thumba.support@thumba.ai